Abstract: A controller for an autonomous vehicle receives audio signals from one or more microphones. The outputs of the microphones are pre-processed to enhance audio features that originated from vehicles. The outputs may also be processed to remove noise. The audio features are input to a machine learning model that classifies the source of the audio features. For example, features may be classified as originating from a vehicle. A direction to a source of the audio features is determined based on relative delays of the audio features in signals from multiple microphones. Where audio features are classified with an above-threshold confidence as originating from a vehicle, collision avoidance is performed with respect to the direction to the source of the audio features.

Abstract: A method for generating training data is disclosed. The method may include executing a simulation process. The simulation process may include traversing a virtual camera through a virtual driving environment comprising at least one virtual precipitation condition and at least one virtual no precipitation condition. During the traversing, the virtual camera may be moved with respect to the virtual driving environment as dictated by a vehicle-motion model modeling motion of a vehicle driving through the virtual driving environment while carrying the virtual camera. Virtual sensor data characterizing the virtual driving environment in both virtual precipitation and virtual no precipitation conditions may be recorded. The virtual sensor data may correspond to what a real sensor would have output had it sensed the virtual driving environment in the real world.

Abstract: Systems, methods, and devices for predicting driver intent and future movements of a human driven vehicles are disclosed herein. A computer implemented method includes receiving an image of a proximal vehicle in a region near a vehicle. The method includes determining a region of the image that contains a driver of the proximal vehicle, wherein determining the region comprises determining based on a location of one or more windows of the proximal vehicle. The method includes processing image data only in the region of the image that contains the driver of the proximal vehicle to detect a driver's body language.

Abstract: Systems, methods, and devices for predicting driver intent and future movements of a human driven vehicles are disclosed herein. A system for predicting future movements of a vehicle includes a camera system, a boundary component, a body language component, and a prediction component. The camera system is configured to capture an image of a vehicle. The boundary component is configured to identify a sub-portion of the image corresponding to an area where a driver of a vehicle is located. The body language component configured to detect a driver's body language. The prediction component configured to predict future motion of the vehicle based on the driver's body language detected by the body language component.

Abstract: A machine learning model is trained by defining a scenario including models of vehicles and a typical driving environment. A model of a subject vehicle is added to the scenario and sensor locations are defined on the subject vehicle. A perception of the scenario by sensors at the sensor locations is simulated. The scenario further includes a model of a parked vehicle with its engine running. The location of the parked vehicle and the simulated outputs of the sensors perceiving the scenario are input to a machine learning algorithm that trains a model to detect the location of the parked vehicle based on the sensor outputs. A vehicle controller then incorporates the machine learning model and estimates the presence and/or location of a parked vehicle with its engine running based on actual sensor outputs input to the machine learning model.

Abstract: Example blind spot detection systems and methods are described. In one implementation, a primary vehicle detects a secondary vehicle ahead of the primary vehicle in an adjacent lane of traffic. A method determines dimensions of the secondary vehicle and estimates a vehicle class associated with the secondary vehicle based on the dimensions of the secondary vehicle. The method also identifies a side-view mirror location on the secondary vehicle and determines a blind spot associated with the secondary vehicle based on the vehicle class and the side-view mirror location.

Abstract: Example blind spot detection systems and methods are described. In one implementation, a primary vehicle detects a secondary vehicle ahead of the primary vehicle in an adjacent lane of traffic. A method determines dimensions of the secondary vehicle and estimates a vehicle class associated with the secondary vehicle based on the dimensions of the secondary vehicle. The method also identifies a side-view mirror location on the secondary vehicle and determines a blind spot associated with the secondary vehicle based on the vehicle class and the side-view mirror location.

Abstract: The present invention extends to methods, systems, and computer program products for distinguishing lane markings for a vehicle to follow. Automated driving or driving assist vehicles utilize sensors to help the vehicle navigate on roadways or in parking areas. The sensors can utilize, for example, the painted surface markings to help guide the vehicle on its path. Aspects of the invention use a first type of sensor and at least a second different type of sensor to identify road surface markings. When ambiguity is detected between road surface markings, decision making algorithms identify the correct set of markings for a vehicle to abide by. The sensors also identify the location and trajectory of neighboring vehicles to increase confidence with respect to the identified road-surface markings.

Abstract: A method is disclosed for using a camera on-board a vehicle to determine whether precipitation is failing near the vehicle. The method may include obtaining multiple images. Each of the multiple images may be known to photographically depict a “rain” or a “no rain” condition. An artificial neural network may be trained on the multiple images. Later, the artificial neural network may analyze one or more images captured by a first camera secured to a first vehicle. Based on that analysis, the artificial neural network may classify the first vehicle as being in “rain” or “no rain” weather.

Abstract: The present invention extends to methods, systems, and computer program products for detecting available parking spaces in a parking environment. Radar systems are utilized to gather data about a parking lot environment. The radar data is provided to a neural network model as an input. Algorithms employing neural networks can be trained to recognize parked vehicles and conflicting data regarding debris, shopping carts, street lamps, traffic signs, pedestrians, etc. The neural network model processes the radar data to estimate parking space boundaries and to approximate the parking space boundaries as splines. The neural network model outputs spline estimations to a vehicle computer system. The vehicle computer system utilizes the spline estimates to detect available parking spaces. The spline estimates are updated as the vehicle navigates the parking environment.

Abstract: The present invention extends to methods, systems, and computer program products for formulating lane level routing plans. In general, aspects of the invention are used in motorized vehicles to guide a driver to a terminal vehicle configuration according to a lane level routing plan that balances travel time with routing plan robustness. A lane level routing plan can be based on terminal guidance conditions (e.g., exiting a highway in the correct off ramp lane), statistical patterns of lanes themselves, current vehicle state, and state of the local environment near the vehicle. Lane level routing plans can be communicated to the driver with audio, visual, and/or haptic cues. Lane level routing plans can be revised online and in (essentially) real-time in response to changing conditions in the local environment (e.g., a trailing vehicle in a neighboring lane has decided to increase speed).

Abstract: A controller for an autonomous vehicle receives audio signals from one or more microphones. The audio signals are input to a machine learning model that classifies the source of the audio features. For example, features may be classified as originating from a vehicle. A direction to a source of the audio features is determined based on relative delays of the audio features in signals from multiple microphones. Where audio features are classified with an above-threshold confidence as originating from a vehicle, collision avoidance is performed with respect to the direction to the source of the audio features. The direction to the source of the audio features may be correlated with vehicle images and/or map data to increase a confidence score that the source of the audio features is a parked vehicle with its engine running. Collision avoidance may then be performed with potential paths of the parked vehicle.

Abstract: The present invention extends to methods, systems, and computer program products for formulating lane level routing plans. In general, aspects of the invention are used in motorized vehicles to guide a driver to a terminal vehicle configuration according to a lane level routing plan that balances travel time with routing plan robustness. A lane level routing plan can be based on terminal guidance conditions (e.g., exiting a highway in the correct off ramp lane), statistical patterns of lanes themselves, current vehicle state, and state of the local environment near the vehicle. Lane level routing plans can be communicated to the driver with audio, visual, and/or haptic cues. Lane level routing plans can be revised online and in (essentially) real-time in response to changing conditions in the local environment (e.g., a trailing vehicle in a neighboring lane has decided to increase speed).

Abstract: A machine learning model is trained by defining a scenario including models of vehicles and a typical driving environment. A model of a subject vehicle is added to the scenario and sensor locations are defined on the subject vehicle. A perception of the scenario by sensors at the sensor locations is simulated. The scenario further includes a model of a lane-splitting vehicle. The location of the lane-splitting vehicle and the simulated outputs of the sensors perceiving the scenario are input to a machine learning algorithm that trains a model to detect the location of a lane-splitting vehicle based on the sensor outputs. A vehicle controller then incorporates the machine learning model and estimates the presence and/or location of a lane-splitting vehicle based on actual sensor outputs input to the machine learning model.

Abstract: Systems, methods, and devices for detecting a vehicle's turn signal status for collision avoidance during lane-switching maneuvers or otherwise. A method includes detecting, at a first vehicle, a presence of a second vehicle in an adjacent lane. The method includes identifying, in an image of the second vehicle, a sub-portion containing a turn signal indicator of the second vehicle. The method includes processing the sub-portion of the image to determine a state of the turn signal indicator. The method also includes notifying a driver or performing a driving maneuver, at the first vehicle, based on the state of the turn signal indicator.

Abstract: The present invention extends to methods, systems, and computer program products for distinguishing lane markings for a vehicle to follow. Automated driving or driving assist vehicles utilize sensors to help the vehicle navigate on roadways or in parking areas. The sensors can utilize, for example, the painted surface markings to help guide the vehicle on its path. Aspects of the invention use a first type of sensor and at least a second different type of sensor to identify road surface markings. When ambiguity is detected between road surface markings, decision making algorithms identify the correct set of markings for a vehicle to abide by. The sensors also identify the location and trajectory of neighboring vehicles to increase confidence with respect to the identified road-surface markings.

Abstract: Systems, methods, and devices for predicting a driver's intention and future movements of a proximal vehicle, whether an automated vehicle or a human driven vehicle, are disclosed herein. A system for predicting future movements of a vehicle includes an intersection component, a camera system, a boundary component, and a prediction component. The intersection component is configured to determine that a parent vehicle is near an intersection. The camera system is configured to capture an image of the proximal vehicle. The boundary component is configured to identify a sub-portion of the image containing a turn signal indicator on the proximal vehicle. The prediction component is configured to predict future movement of the proximal vehicle through the intersection based on a state of the turn signal indicator.

Abstract: Systems, methods, and devices for predicting driver intent and future movements of a human driven vehicles are disclosed herein. A system for predicting future movements of a vehicle includes a camera system, a boundary component, a body language component, and a prediction component. The camera system is configured to capture an image of a vehicle. The boundary component is configured to identify a sub-portion of the image corresponding to an area where a driver of a vehicle is located. The body language component configured to detect a driver's body language. The prediction component configured to predict future motion of the vehicle based on the driver's body language detected by the body language component.

Abstract: Methods and apparatus pertaining to an intelligent vehicle access point opening system are provided. A method may involve detecting a presence of an object in a vicinity of a cover of an access point of a vehicle. The method may also involve receiving a command to open the cover and activating a mechanism to open the cover responsive to receiving the command. The method may further involve determining whether the object is likely to fall as the cover is being opened. The method may additionally involve pausing opening of the cover responsive to a determination that the object is likely to fall as the cover is being opened.

Abstract: A controller for an autonomous vehicle receives audio signals from one or more microphones. The outputs of the microphones are pre-processed to enhance audio features that originated from vehicles. The outputs may also be processed to remove noise. The audio features are input to a machine learning model that classifies the source of the audio features. For example, features may be classified as originating from a vehicle. A direction to a source of the audio features is determined based on relative delays of the audio features in signals from multiple microphones. Where audio features are classified with an above-threshold confidence as originating from a vehicle, collision avoidance is performed with respect to the direction to the source of the audio features.